Optimal Periodic Sensor Scheduling in Large-Scale Dynamical Networks
Sijia Liu, Makan Fardad, Engin Maşazade, Pramod K. Varshney · 2013
We consider the problem of finding optimal time-periodic sensor schedules for estimating the state of a large-scale dynamical system. We assume that a large number of sensors have been deployed and that the sensors are subject to resource constraints, which limits the number of times each can be activated over one period. We seek an algorithm that strikes a balance between estimation accuracy and total sensor activations over one period. We make a correspondence between active sensors and the nonzero columns of the estimator gain, and formulate an optimization problem in which the estimator gain minimizes the trace of the error covariance while being penalized for its number of nonzero columns. This optimization problem is combinatorial in nature, and we employ the alternating direction method of multipliers (ADMM) to find its locally optimal solutions. Numerical results are provided to illustrate the effectiveness of our proposed method. Index Terms Dynamic system, state estimation, sensor scheduling, sparsity, sensor networks.